Abstract

The criterion of error of misclassification of Fisher linear classifier has not been fully investigated in the context of intra class correlations in classifying two Multivariate Normal populations [Johnson and Wichern (2001)]. Here an attempt is made to study the performance of Fisher classifier in classifying a new observation X 0 into one of the two multivariate normal populations N p (µ1, S) and N p (µ2 , S) in the above said context using simulation study.

Author: A.V. Dattatreya Rao and Nunna Srinivasa Rao

Received on: July, 2004

Accepted on: